The current information technology revolution is far-reaching, and the automotive industry is undergoing unprecedented changes in a century. People's traditional perceptions of cars are being subverted. Smart cockpits and smart driving are moving from concept to product. The product attributes, industrial value chains and ecological structures of automobiles are being reconstructed through the power of science and technology and imagination. Under the wave of automotive intelligence, how can Chinese manufacturers seize the opportunity to surpass in the corners?
In order to help auto companies and employees better grasp the development trend of China's smart cars, understand consumer needs and willingness to pay for emerging car consumption scenarios, Autohome Research Institute and 21 New Automotive Research Institute released the "2022 China Intelligent Vehicle Development Trend Insight Report" (hereinafter referred to as the report). The report focuses on users, products, and businesses. Based on the survey samples of more than 20,000 real users, the clue data of more than 100 million models, the configuration data of more than 60,000 models, the professional evaluation data of 18 mainstream new cars, and the interviews and research of more than 10 core companies in the industry, this report explores the development trend of future products, mines user needs and expectations through massive data, points out problems and proposes solutions.
【Auto Home Research Institute】Auto Home Research Institute was officially established in March this year, aiming to build a forward-looking research think tank for the automotive industry, using industry advantages to provide advice on a series of important issues in the development of China's automotive industry. Autohome Research Institute brings together the research strength of top industry experts at home and abroad, using the full-link data resources of viewing, selecting, purchasing and using, as well as years of industry observation and precipitation, to gain insight into the new trends of the automotive industry, new changes in users and new competitive patterns to enhance the ability of the automotive industry. At present, the institute has five research centers: the Automotive Industry Research Center, the New Energy Research Center, the Used Car Research Center, the Aftermarket and Ecology Research Center, and the Consumer Research Center.
【21st Century New Automobile Research Institute】In order to integrate and strengthen resources, expand and strengthen its professionalism and influence in the entire automotive industry, 21st Century Business Herald established the 21st Century New Automobile Research Institute in November 2021, focusing on the new trend of automobile development in the era of intelligent electric vehicles, studying major topics such as the reconstruction of the automotive industry chain from the production end to the consumer end and even the entire automotive industry chain, and producing and publishing relevant research results and reports. It is committed to promoting the effective integration and utilization of resources in the entire automotive industry chain, and promoting and shaping a better ecosystem for the entire automotive industry chain.
Based on the overall analysis, the report draws the following important judgments:
First of all, China's smart car market has great potential. Relevant policies have shifted from setting goals to guiding landing and infrastructure construction, and cooperation between technology companies and traditional automobile companies has been strengthened;
Second, the market penetration rate of intelligent cockpit configuration and intelligent driving configuration is growing rapidly, and intelligent feature configuration will become a new weapon in the differentiated market competition of Chinese brands;
Third, the IV rating test scores of Chinese brand models are eye-catching;
Fourth, young people and highly educated people have a higher acceptance of automotive intelligence;
Fifth, 5G technology promotes the development of the v2x market, and the coordinated development of vehicle road coordination and smart bicycles will become the main theme.
The development status of intelligent vehicles: broad market space, policy support, infrastructure, advantages, win-win cooperation
It is expected that the smart cockpit market size will exceed 100 billion in 2025. According to icvtank, the global smart cockpit industry market is expected to reach $46.1 billion by 2022. As the world's most promising automotive market, China's smart cockpit market will reach 44.1 billion yuan in 2019 and 103 billion yuan in 2025.
The penetration rate of smart cars has increased rapidly, and China's smart car market has great potential. With the continuous downward exploration of the cost of key components, the maturity of the superimposed industrial environment and the continuous progress of science and technology, it is expected that by 2025, the sales volume of smart cars in China will exceed 10 million, and the corresponding penetration rate of smart cars will reach 49.3%.
Smart cars involve multiple industries, such as big data, artificial intelligence, cloud computing, semiconductors, information and communications and other high-end manufacturing fields. They have become "battlefields" for the world's major countries. China's policies have gradually shifted from setting goals and setting norms to guiding implementation, building infrastructure, and cyber and data security.
Autonomous driving technology companies basically take L4 and above automatic driving as the research and development goal. At present, China's leading enterprises mainly provide a full range of solutions. Two commercialization models have been realized, one is autonomous driving to solve simple scenarios, and the other is that traditional auto companies and technology companies cooperate through cooperation or mergers and acquisitions.
At present, many collaborations are done to complete the optimization of algorithms and the integration of hardware and software. With the intensification of competition and the slowdown in the pace of financing, auto companies may make up for their shortcomings in the field of autonomous driving through mergers and acquisitions.
"One core and multiple screens" provides unlimited imagination space for the development of intelligent cockpits. The smart cockpit is more easily perceived by the user, and this technology is less difficult to develop than intelligent driving. The intelligent cockpit is connected to the owner's life ecology, and the use scenarios will be infinitely expanded.
The hardware configuration of the new power brand is advanced, and the pressure of algorithm optimization and iterative competition is great. At present, the hardware configuration of China Xinli can basically meet the needs of the current stage and the upgrading of the owner's automatic driving service in the next few years. However, manufacturers still need to continuously optimize and iterate on perception algorithms, planning algorithms, and control algorithms, and integrate the execution and interconnection layers to achieve higher-order autonomous driving.
Intelligent configuration development trend: Rapid penetration of Internet connection functions, gradual popularization of L2 level automatic driving, and featured configurations to help differentiate competition
In terms of smart cockpits, interactive configurations are trying to break the "cycle". The penetration rate of the network connection function configuration market has risen rapidly, with an average five-year compound growth rate of nearly 60%. Among them, the five-year compound growth rate of Ota upgrade is as high as 122.7%. Rapid iterative updates could become essential equipment for smart vehicles. However, the market penetration of human-computer interaction configuration is not high, and 5G and car networking technology may help it break the "cycle" in the future.
In addition, the report also analyzes in detail the market penetration trends of different brand grades (overseas luxury brands, Chinese brands and overseas mainstream brands) such as "Internet of Vehicles", "Navigation Road Condition Information Display", and "interactive functions" such as in-car mood lights and mobile phone wireless charging.
In terms of smart driving configurations, the average compound growth rate over the past five years has reached 42.7%. At present, the market penetration rate of the primary drive core configuration is growing; the full-speed adaptive cruise is growing the fastest; and the automatic parking has entered the platform stage, and the penetration rate is slowly increasing.
The market penetration of L2 models has grown rapidly, up 24.5 times compared to 2017. As autopilot hardware prices continue to fall, L2-level configurations have penetrated into the low-cost segment. By 2021, 200,000-300,000 models have entered the first camp, with a penetration rate of 37.6%. By 2021, the market penetration rate of new energy vehicle L2 autonomous driving will reach 33.7%, much higher than the 19.0% of non-new energy vehicles.
In addition, intelligent function configuration is the selling point of differentiated competition between brands. Chinese brands started earlier and developed faster in this regard. The differential advantages of these emerging configurations could become a new weapon for Chinese brands to compete in the market.
The current situation of smart car level: Chinese brands have performed eye-catchingly, but there is a lot of room for product improvement, and safe driving needs to be improved
The average score of all brands in the IV rating test is 66.4 points. Among them, the performance of Chinese brand models is eye-catching. Iv Rating is a "smart car evaluation system" designed by Autohome. Professional evaluation is carried out from the two dimensions of intelligent cockpit and intelligent driving, including 8 dimensions of the user's actual use scenario. The test strictly followed 36 rigorous criteria and 75 subdivision evaluation criteria.
Overall, there is still a lot of room for improvement in smart car products. By the end of 2021, the average total score of the 18 mainstream smart cars participating in the test was only 66.4 points; the proportion of vehicles with scores below 65 points was as high as 44.5%.
The pros and cons are obvious. The development of smart cockpits and smart driving is uneven. The main shortcomings of the smart cockpit consist of two parts: scene coverage and communication preferences; the main weakness of smart driving lies in the high speed and parking lot view. Overall, the performance of the Chinese brand car series is more eye-catching, with the total score of the top three coming from Chinese brands.
In the top ten smart cockpit scores in 2021, Chinese brand cars occupied 8 seats. Among them, the Haval H6 and Hoshigoku L belong to the first camp, and their performance is significantly higher than other models.
From the perspective of various dimensions, Xiaopeng P7 has the highest interactive convenience score, and the voice function is particularly powerful; the ideal one has the highest function satisfaction score, which has various entertainment systems; Xingyue L has the highest scene coverage score, which is due to the fact that the mini program embedded in the system covers many automobile and life scenes; and Weilai ec6 scores the highest score in communication support. The NOMI robot adds a warm interactive experience to human-machine interaction.
In many aspects of the smart cockpit, Chinese brands are in the lead. Chinese brands scored the lowest on the mapping feature. Some of the new smart cars introduced in recent years do not support any type of mobile map feature. They argue that the cockpit is enough to replace or even surpass mapping tools like Carplay or Carlife. Luxury brands are weaker than other camps in terms of scene coverage, and obviously lack localized service ecology.
In terms of intelligent driving, tesla, the benchmark of self-driving in the traditional impression, is not as good as Chinese brands, ranking only eighth. In the top ten, all models except the third Mercedes-Benz E-Class and the eighth Tesla Model y are Chinese brands.
The parking lot view is unique to Chinese brands, but the scene related to safe driving needs to be improved. Chinese brand models are particularly prominent in the parking lot test, scoring much higher than other camps. Among them, the basic ability and parking ability of various types of parking spaces are obviously stronger, but the intelligent driving ability of safety-related scenarios needs to be improved.
The implementation of AEB functions is mainly based on hardware fusion schemes, but the performance varies greatly from training method to training method. At present, the path selection of perception technology is still based on the hardware fusion scheme of vision + radar. However, the choice of scheme does not completely determine the effect of AEB, and it also needs high-quality chips and advanced algorithms to support it to better adjust the performance of the model.
Assisted driving is primarily monocular driving, but advanced autopilot lidar solutions are essential. In the iv-level test, most of the vehicles with good performance use the technical solution of multi-camera + angle radar, and the multi-hardware solution may not be redundant in high-speed scenarios.
Analysis of user demand preferences: Young and highly educated groups are more open-minded, and post-00s are more conservative
In terms of accepting autonomous driving, 40% of users prefer to accept assisted driving. From generation to generation, except after the age of 00, the younger you are, the more willing you are to accept advanced autonomous driving. From the perspective of car purchase budget, the higher the budget, the higher the acceptance of advanced autonomous driving. In terms of educational background, the higher the educational background, the higher the acceptance of advanced autonomous driving.
In terms of traditional smart car configuration functions, the new generation group does not pay. From a generational perspective, people have similar preferences for the first six configurations of traditional cars, but they are not interested in the configuration of the post-00s. Active braking/active safety systems and full-speed adaptive cruise are the two most valued configurations for the post-80s and post-90s, and the proportion of post-00s choosing them is much lower. The traditional configuration seems to be difficult to impress the new generation of post-00s.
In addition, the report also analyzes the preferences of different groups of people for various traditional smart vehicle configurations, such as active braking, full-speed adaptive cruise, lane merge assist, lane departure warning systems and seat ventilation, as well as the preferences of different groups of people for various emerging intelligent configurations, such as mobile application remote control vehicles, A-pillar/chassis perspective image systems, and AR real-time map navigation.
For driverless public transport, young people and highly educated users are more willing to pay. Overall, more than 60 percent of users are willing to accept driverless public transportation, with the highest acceptance rate after age 70; after 20:00, this is unacceptable due to safety concerns. The younger the user, the more they are receptive to advanced driverless buses; the lower the level of education, the lower the price of driverless buses is lower than that of ordinary buses.
v2x status quo and commercialization: intelligent high speed, long payback period, large-scale configuration of robotaxi, terminal logistics cost reduction space
According to data released by ICVTANK, it is expected that by 2022, the global v2x market size will exceed 165 billion US dollars, and the Chinese v2x market size will reach 50 billion US dollars. At present, the vehicle terminal OBU products are mainly T-box, with a price of about 2,000 yuan. It is estimated that by 2023, the front-end configuration of T-box boxes for passenger cars in China will reach 14 million units.
In the future, the coordinated development of vehicle roads and smart bicycles will become the main theme of the development of intelligent cars. Vehicle-road collaborative autonomous driving has become a clear technical route for the mainland to develop high-level automatic driving. The higher the level of autonomous vehicle driving, the lower the requirements on the road. It has built nearly 4,000 kilometers of smart highways with the government and 30 intelligent highways with the government.
Bicycle intelligence and vehicle road coordination are more like the "right arm" of intelligent driving. In the case of complex intersections, for example, drivers and sensors are limited by viewing angles and line of sight, while roadside sensing devices and computing platforms can sense all traffic participants within the intersection. Vehicle-to-road coordination has become a necessary condition for the development of autonomous driving. According to the prediction of the Future Industry Research Institute, by 2030, the cumulative investment scale of major IT equipment of Chelu Cooperation will reach 283.4 billion yuan.
The report also calculates the cost-effectiveness of three business models for applying v2x. Through the analysis and comparison of Meituan's automatic vehicle allocation and driver's car allocation, it is believed that after the implementation of Meituan's automatic vehicle allocation, the expenditure cost can be reduced by half; the investment payback period of the Hangzhou-Shaoxing-Ningbo intelligent highway project is evaluated and calculated. It is believed that there are problems in the acceptance of high-intelligence highway toll users and a long project recovery period; through the benchmarking analysis of the cost of automatic driving and the cost of driver driving, it is believed that the robotaxi market is huge, but the platform needs to reduce the cost of early research and development through scale effect.
Smart car development suggestions: adjust the cockpit configuration, increase the L2 standard configuration rate, and full-stack independent development strategy
In terms of smart cockpit configuration, most of the configuration items are in short supply, which reflects the willingness of users to pay more for the purchase cost of the car and buy a higher configuration model. 300,000-500,000 models in particular need to increase the standard configuration rate of network connection configurations.
In terms of intelligent driving configuration, L2 assisted driving related configurations and safety configurations are deeply loved by consumers. Increasing the standard configuration rate of L2 driver assistance configurations is conducive to market competition. The panoramic satellite navigation system is in a state of simultaneous power supply and automatic parking line.
Finally, from the perspective of problem solving and future development direction, the report makes six suggestions for Chinese automakers: (1) OTAs should not be reduced to marketing or patching tools, and OEMs should take back the initiative as soon as possible; (2) in the early stage of autonomous driving development, safety and security should be the primary consideration; (3) "taking the way" has obvious short-term benefits, but a full set of self-learning must be the ultimate attribution; (4) open source, innovation and breakthrough, to establish systematic and strong software capabilities; (5) AEB and mobile remote control vehicles have the highest willingness to pay, and can give priority to the purchase of water testing software; (6) improve the acceptance of intelligent driving users through intelligent driving activities, achievement displays, functional experience, etc.
The information technology revolution we are in is profoundly changing the automotive industry. China's auto industry should seize the opportunities of the digital and intelligent era and strive to catch up and surpass. The marathon for the development of China's autonomous driving technology is halfway through. L2 self-driving cars are in the commercialization and development phase. V2x has also begun market exploration. Automotive intelligence opens up business model innovations, but they need to be explored and validated.